Depression is among the most prevalent of all psychiatric disorders, accounting for over 20% of economic costs for all mental illness. An important public health priority is the development of methods for identifying factors that both increase individuals' vulnerability to depression and hinder recovery from this disorder. Pursuing this objective, we have been conducting research that broadens the focus of cognitive models of depression by examining patterns of biological reactivity to emotional stimuli. Extrapolating from our earlier studies and from existing models, we predicted that individuals diagnosed with Major Depressive Disorder (MDD) would exhibit increased autonomic arousal and neural activation in response to negative stimuli. But our findings indicated otherwise; in both fMRI and psychophysiological studies, MDD individuals demonstrated less activation and arousal in response to both positive and negative stimuli than did nondepressed controls. Moreover, greater activation and reactivity predicted improved functioning over the following year, independent of MDD severity. In reconciling and integrating these results with our earlier findings, we are broadening and refining cognitive theories of depression to include a consideration of affective chronometry, sensitivity to reward stimuli, and biological functioning, not only to describe the functioning of depressed individuals, but also to predict recovery from this disorder. The studies proposed in this application are designed to test a more comprehensive formulation of depression and to systematically rule out important alternative explanations for our obtained results. Based on our initial findings, we posit that when confronted with an emotionally-valenced stimulus, depressed persons immediately attend to it; if the stimulus persists, they shut down their processing and disengage from it. Moreover, those depressed individuals who are able to stay engaged with emotional stimuli are likely to recover from depression most quickly. The overall objective of this proposed project is to advance our understanding of the cognitive and neurobiological responses in MDD to emotional stimuli. We are particularly interested in the effects of stimulus type (i.e., affective valence), stimulus duration (i.e., affective chronometry), and the potential effects of specific patient characteristics (e.g., anhedonia, comorbid Axis-II personality disorder) on emotional functioning in depression. We propose to examine information-processing biases and reward responsivity, as well as concomitant neural activation and psychophysiological arousal, in response to emotional stimuli in depressed and nondepressed participants. Findings from the proposed studies testing these formulations will contribute importantly to the development of an integrative psychobiological theory of depression.